Project Vesta, a major experiment investigating the behaviour of high-intensity fires with the aim of developing more accurate ways to predict bushfire spread, is set to continue in Western Australia this summer.
The experiment is being undertaken jointly by CSIRO and the Western Australian Department of Conservation and Land Management (CALM).
It involves detailed study of hot fires lit in two areas of jarrah forest in the State's south-west of Western Australia - one, near Margaret River, with relatively high rainfall and the other in drier country east of Harvey.
According to Phil Cheney, leader of the CSIRO Forestry and Forest Products Fire Research Group, some important findings have emerged already - including strong indications that the relationships between fuel load, wind speed and fire spread assumed in current prediction systems need to be reassessed. `In moderate to medium-intensity wildfires, some things are happening that we are not predicting at all well,' he says.
Particularly significant, not least for firefighter safety, are results suggesting that the direct relationship assumed between fuel load and rate of fire spread - that a doubling of fuel doubles the rate of spread - is far from universally applicable. `We had some fires that showed a much steeper direct relationship, with a doubling of fuel load giving a 5--6 times increase in the rate of spread,' Cheney says. `Another set of fires showed no consistent relationship.'
In some cases heavy fuel appeared to actually reduce the rate of spread. Cheney suspects that, at low wind speeds, strong convection from these fires draws the flames inward, tending to make a fire's progress slower and more erratic. But when a threshold wind speed, which varies with the fuel load, is exceeded the situation changes dramatically, with the fire rapidly achieving its potential rate of spread. He thinks that under some conditions this may be as much as 10 times greater than the rate that would be predicted using current methods.
Data from this summer's fires will add more precision to these findings. Another conclusion that they will throw further light on is that, like grassfires, forest fires have to be quite wide before they reach their maximum rate of spread.
Under high wind conditions, fire speed appears to increase fairly rapidly with width up to a front length of about 100 metres, and then more slowly until the maximum is reached at around 250 metres.
An important finding from wind measurements during last summer's research is that gusts tend not to travel far in a forest, with the result that readings taken more than about 40 metres downwind from a fire may have little predictive value. `It appears that a gust flows a small distance, dissipates, and is replaced by another gust,' Cheney says.
`You can get quite big changes of wind speed in a forest that have a relatively localized effect. The danger is that if a gust exceeds a fire's threshold wind speed the fire might suddenly take off at a rate 3--5 times faster than before.'
Information gained from Project Vesta will be critical to the ongoing development of systems used to predict fire behaviour and assess the threats posed by uncontrolled fires, according to CALM Senior Research Scientist Lachie McCaw.